Fake News Detection

• 

  Model Accuracy, 0.93

• 

  Random News Detection

• 

  Detect News Content through News Link

• 

  Detect News Content through Pasted Content

Inspiration

I got inspired by ScamShield. ScamShield has features to block scam messages and spam calls. I thought of having another real-life problem that needs AI detection like ScamShield, such as reading the news. When we read news online, we do not which is the truth, or which is fake. Therefore, I have come up with the idea of having a Fake News Detection tool.

What it does

As the name mentions, it helps detect whether news is fake or not. Additional features include showcasing the confidence level, and having the option to insert news via URL/text.

How I built it

Tool used: Jupyter Notebook Dataset used: hf://datasets/ErfanMoosaviMonazzah/fake-news-detection-dataset-English

Challenges I ran into

Having to build and train an model from scratch, without any prior AI experience was tedious. It took me a long time to figure out how to code it.

In addition, I tried to build a website linking with the model but it does not run smoothly.

Accomplishments that I proud of

I proud that I manage to train this model individually, within a day.

What I learned

I learn to not give up easily and enhance on the model in the future.

What's next for Fake News Detection

Hmm… Let's find out!

Built With

• jupyter
• pandas
• python
• sklearn